Establishing 99.9% availability for Partner Interconnect

In this tutorial, you use Partner Interconnect to create and configure resources to achieve 99.9% availability. This topology is suitable for non-critical applications that can tolerate some downtime. For example, your organization might do batch uploads for which the timing of the operation isn't critical. If a connection isn't available, the operation continues when connectivity is restored.

For Cloud Interconnect terminology used on this page, see Key terms.

For information about creating redundant Interconnect connections that have sufficient capacity in a failover scenario, see Best practices.

Requirements

The following resources and settings are required to achieve 99.9% availability:

  • At least two VLAN attachments in a single Google Cloud region, in separate edge availability domains (metro availability zones), each with their own Cloud Router.

  • The attachments must connect in one metropolitan area (metro). For more information about choosing an Interconnect connection location, see Requesting connections.

  • At least two Cloud Routers, each attached to a VLAN attachment in a separate edge availability domain.

  • Depending on your hardware and availability requirements, you might have one or more routers in your on-premises network.

Example topology

This example topology uses the following resource names, regions, and Interconnect connection locations. As you walk through the tutorial, replace these values with your own.

Region Router name VLAN attachment name Connection location Edge availability domain
us-central1 router-central-a attach-central-a Dallas 1
us-central1 router-central-b attach-central-b Dallas 2

The following diagram shows two layer 2 connections in two different edge availability domains in a single metro: Dallas - domain 1 and Dallas - domain 2.

The Cloud Routers (one for each VLAN attachment) are in the vpc1 network in the us-central1 region. For layer 2, users configure Border Gateway Protocol (BGP) sessions on their on-premises routers to each of their Cloud Routers.

Redundant connections for 99.9% availability using layer 2 (click to enlarge).
Redundant connections for 99.9% availability using layer 2 (click to enlarge)

Similar to the layer 2 topology, the following layer 3 topology shows two Interconnect connections in two different edge availability domains in a single metro. The Cloud Routers (one for each VLAN attachment) are in the vpc1 network in the us-central1 region. For layer 3, the service provider configures the BGP sessions.

Redundant connections for 99.9% availability using layer 3 (click to enlarge).
Redundant connections for 99.9% availability using layer 3 (click to enlarge)

Creating a 99.9% topology

The following sections walk through the steps for creating this topology. To create a topology that's tailored for you, replace the sample input values with your own. For example, replace the vpc1 network name with your Virtual Private Cloud (VPC) network name.

Create the VPC network

For the vpc1 network, create a network with custom subnets and set its dynamic routing mode to global. For more information, see Creating a custom mode network in the VPC documentation.

Console

  1. In the Google Cloud Console, go to the VPC networks page.

    Go to VPC networks

  2. Click Create VPC network.

  3. For Name, enter vpc1.

  4. For Subnet creation mode, select Custom.

  5. Specify the details of your subnet:

    1. Name: Enter subnet-uscentral1.
    2. Region: Select us-central1.
    3. IP address range: Specify an IP range, such as 192.168.1.0/24. Use a range that doesn't overlap with your on-premises network.
  6. Click Add subnet.

  7. Specify the details of your second subnet:

    1. Name: Enter subnet-useast4.
    2. Region: Select us-east4.
    3. IP address range: Specify an IP range, such as 192.168.2.0/24. Use a range that doesn't overlap with your on-premises network.
  8. In the Dynamic routing mode section, select Global.

  9. Click Create.

gcloud

  1. Create a custom subnet:

    gcloud compute networks create vpc1 \
       --subnet-mode custom \
       --bgp-routing-mode global
  2. Specify the subnet prefixes for the us-central1 and us-east4 regions:

    gcloud compute networks subnets create subnet-uscentral1 \
       --network vpc1 \
       --region us-central1 \
       --range 192.168.1.0/24
    gcloud compute networks subnets create subnet-useast4 \
       --network vpc1 \
       --region us-east4 \
       --range 192.168.2.0/24

Create Cloud Routers

Create two Cloud Routers, one for each VLAN attachment. For Partner Interconnect, all Cloud Routers must have a local ASN of 16550.

Console

  1. In the Cloud Console, go to the Create a Cloud Router page.

    Go to Create a Cloud Router

  2. Specify the Cloud Router details:

    • Name: Enter router-central-a for the name of the first Cloud Router.
    • Network: Select vpc1, which is the name of the network where the on-premises network extends to.
    • Region: Select us-central1, which is the region where the Cloud Router is configured, as shown in the example topology.
    • Google ASN: Enter 16550, which is required for Partner Interconnect.
  3. Under Routes, select Advertise all subnets visible to the Cloud Router (default), and then click Create.

  4. Follow the same steps to create another Cloud Router, but use the name router-central-b.

gcloud

  • Create two Cloud Routers in the vpc1 network in the us-central1 region; use ASN 16550 for both Cloud Routers:

    gcloud compute routers create router-central-a \
      --asn 16550 \
      --network vpc1 \
      --region us-central1
    
    gcloud compute routers create router-central-b \
      --asn 16550 \
      --network vpc1 \
      --region us-central1
    

Create VLAN attachments

Create two VLAN attachments and pair them with Cloud Routers, as shown in the following list:

  • router-central-a is associated with attach-central-a
  • router-central-b is associated with attach-central-b

Console

  1. In the Cloud Console, go to the Interconnect Hybrid Connectivity page.

    Go to Interconnect Hybrid Connectivity

  2. Click Get started.

  3. Select Partner Interconnect, and then click Continue.

  4. Select I already have a service provider.

  5. Select Create a redundant pair of VLAN attachments (recommended).

  6. For the Network and Region fields, select the vpc1 network and the us-central1 region, which specifies the VPC network and region to connect to.

  7. Specify the details of the first VLAN attachment (VLAN A):

    • Cloud Router: Select router-central-a.
    • VLAN attachment name: Enter attach-central-a. This name is displayed in the Cloud Console and is used by the gcloud command-line tool to reference the attachment.
  8. Specify the details of the second VLAN attachment (VLAN B):

    • Cloud Router: Select router-central-b.
    • VLAN attachment name: Enter attach-central-b.
  9. Click Create to create the attachments, which takes a few moments to complete.

  10. Click OK to view a list of your VLAN attachments.

gcloud

  • Create two attachments:

    gcloud compute interconnects attachments partner create attach-central-a \
       --router router-central-a \
       --region us-central1 \
       --edge-availability-domain availability-domain-1
    gcloud compute interconnects attachments partner create attach-central-b \
       --router router-central-b \
       --region us-central1 \
       --edge-availability-domain availability-domain-2

Retrieve pairing keys

After you create your VLAN attachments, Google Cloud generates pairing keys for each one. Pairing keys enable service providers to identify your VLAN attachment without you having to share information about your project or resource names.

Console

  1. In the Cloud Console, go to the VLAN attachments tab.

    Go to VLAN attachments

  2. Find the attach-central-a VLAN attachment, and then click View instructions.

    The Cloud Console shows the pairing key for the attachment. Copy and record the key; you'll share these keys with your service provider when you request a connection with them.

  3. Click OK to return to the list of VLAN attachments.

  4. Retrieve the pairing keys for the attach-central-b VLAN attachment.

gcloud

  1. Retrieve the VLAN attachment pairing keys by describing each one; use these keys to request connections with your service provider:

    gcloud compute interconnects attachments describe attach-central-a \
       --region us-central1

    The previous command outputs the following information:

    adminEnabled: false
    edgeAvailabilityDomain: AVAILABILITY_DOMAIN_1
    creationTimestamp: '2017-12-01T08:29:09.886-08:00'
    id: '7976913826166357434'
    kind: compute#interconnectAttachment
    labelFingerprint: 42WmSpB8rSM=
    name: attach-central-a
    pairingKey: 7e51371e-72a3-40b5-b844-2e3efefaee59/us-central1/1
    region: https://www.googleapis.com/compute/v1/projects/customer-project/regions/us-central1
    router: https://www.googleapis.com/compute/v1/projects/customer-project/regions/us-central1/routers/router-central-a
    selfLink: https://www.googleapis.com/compute/v1/projects/customer-project/regions/us-central1/interconnectAttachments/attach-central-a
    state: PENDING_PARTNER
    type: PARTNER
    
  2. Record the value of the pairingKey for each VLAN attachment.

Request connections from your service provider

Contact your service provider and request two connections, one for each VLAN attachment. The service provider requires the pairing key, capacity, and location for each attachment.

The location of each attachment must be in a single location (city). In this example, attach-central-a and attach-central-b are connected to Dallas.

After the connection configuration is complete, you receive an email notification indicating that you can activate your attachments.

Activate VLAN attachments

After your service provider configures your VLAN attachments, you must activate them before they can start passing traffic. If you pre-activated your attachments, you can skip this step.

Console

  1. In the Cloud Console, go to the VLAN attachments tab.

    Go to VLAN attachments

  2. Select the attach-central-a VLAN attachment to view its details page.

  3. If the status of your attachment is Waiting for service provider, wait until your service provider has completed the VLAN configuration.

  4. Review the Interconnect information to check that the expected service provider configured the VLAN attachment.

  5. Click Activate to activate the attachment.

  6. Go back to the VLAN attachments tab and activate the other attachment.

gcloud

  1. Describe each VLAN attachment to verify whether your service provider completed configuring them:

    gcloud compute interconnects attachments describe attach-central-a \
       --region us-central1 \
       --router router-central-a \
       --format '(name,state,partnerMetadata)'
    gcloud compute interconnects attachments describe attach-central-b \
       --region us-central1 \
       --router router-central-b \
       --format '(name,state,partnerMetadata)'

    After your service provider configures your VLAN attachments, the state of your attachments changes from PENDING_PARTNER to PENDING_CUSTOMER. If the state is still PENDING_PARTNER, wait until your service provider has completed the VLAN configuration.

  2. Check the partnerMetadata field to verify that the correct service provider configured your VLAN attachments.

    If an unexpected third party configured your VLAN attachment, delete the attachment and then create a new one, which generates a new pairing key. Use the new pairing key to request another connection with your service provider.

  3. If the correct service provider has configured your VLAN attachments, activate them by using the --adminEnabled flag:

    gcloud compute interconnects attachments partner update attach-central-a \
       --region us-central1 \
       --admin-enabled
    
    gcloud compute interconnects attachments partner update attach-central-b \
       --region us-central1 \
       --admin-enabled
    

Configure routers

Google automatically adds a BGP peer on each Cloud Router. For layer 2 connections, you must add your on-premises ASN to each BGP peer. For layer 3 connections, you don't need to do anything; Google automatically configures your Cloud Routers with your service provider's ASN.

Console

  1. In the Cloud Console, go to the VLAN attachments tab.

    Go to VLAN attachments

  2. Select the attach-central-a VLAN attachment to view its details page.

  3. Click Configure BGP to configure the Cloud Router that's associated with this VLAN attachment.

  4. In the Peer ASN field, add the ASN of your on-premises router.

  5. Click Save and continue.

  6. Go back to the VLAN attachments tab and add your on-premises ASN to the other Cloud Routers.

gcloud

  1. Describe the Cloud Router that's associated with the attach-central-a VLAN attachment. In the output, find the name of the automatically created BGP peer that's associated with your VLAN attachment. The values of ipAddress and peerIpAddress of the BGP peer must match the values of cloudRouterIpAddress and customerRouterIpAddress of your attachment.

    gcloud compute routers describe router-central-a \
       --region us-central1
    bgp:
      advertiseMode: DEFAULT
      asn: 16550
    bgpPeers:
    — interfaceName: auto-ia-if-attachment-central-a-c2c53a710bd6c2e
      ipAddress: 169.254.67.201
      managementType: MANAGED_BY_ATTACHMENT
      name: auto-ia-bgp-attachment-central-a-c2c53a710bd6c2e
      peerIpAddress: 169.254.67.202
    creationTimestamp: '2018-01-25T07:14:43.068-08:00'
    description: 'test'
    id: '4370996577373014668'
    interfaces:
    — ipRange: 169.254.67.201/29
      linkedInterconnectAttachment: https://www.googleapis.com/compute/alpha/projects/customer-project/regions/us-central1/interconnectAttachments/customer-attachment1-partner
      managementType: MANAGED_BY_ATTACHMENT
      name: auto-ia-if-attachment-central-a-c2c53a710bd6c2e
    kind: compute#router
    name: partner
    network: https://www.googleapis.com/compute/v1/projects/customer-project/global/networks/vpc1
    region: https://www.googleapis.com/compute/v1/projects/customer-project/regions/us-central1
    selfLink: https://www.googleapis.com/compute/v1/projects/customer-project/regions/us-central1/routers/router-central-a
  2. Update the BGP peer with your on-premises router's ASN:

    gcloud compute routers update-bgp-peer router-central-a \
       --peer-name auto-ia-bgp-attachment-central-a-c2c53a710bd6c2e \
       --peer-asn [ON-PREM ASN] \
       --region us-central1

After the BGP sessions are established, you can pass between your VPC network and on-premises network through your service provider's network.

What's next